PRIMULA. 



insuperable. In this case, rosettes should be taken off and struck as 

 cuttings in sand about the end of July. They will immediately proceed 

 to root, and at the end of August, with two or three young tentacles 

 just beginning to push, can be easily inserted into the crevices that they 

 are to occupy for ever. Thus it was that my original fragments on 

 the Cliff were established there without difficulty some four years since, 

 and now, in the darkest hours of spring till far on into April, their 

 wonderful royal flowers come forth, lying flat upon the tiny tufts, and 

 glorify the grey walls of rock against which they shine like living 

 jewels. In fact, in any circumstance, I prefer August for the planting 

 of all the more tricky Primulas ; as being the moment when the old 

 roots are stationary, and just before the new ones issue forth to anchor 

 the plant in its new situation. In size and shape of blossom P. Allionii, 

 like all its race, is extremely variable, some flowers being of aniline 

 colouring and thin shape, whereas the best are of the most beautiful 

 amplitude and rotundity, and of the blandest softest rose-pink, very 

 often diversified with a clear white eye. P. All'odii can easily be 

 raised from seed, but the quicker and surer method of multiplying it 

 is by offshoots and cuttings taken off in July or August. It is not, 

 I think, patient of pot culture for many seasons together (unless 

 kept on the move), though an exquisite object in pans in a cool 

 greenhouse ; in ordinary circumstances the trouble is to keep the 

 soil in the pot from growing stale and stagnant. In the open, 

 however, with sufficiency (but no excess) of moisture, and in some 

 firm, deep, perfectly-drained and downward-draining limestone 

 crack or crevice, as small and tight as possible, P. Allionii, whether 

 in sun or shade, is well capable of forming the annually increas- 

 ing glory of the rock-garden, now that the pernicious days of 

 pockets and fussments have passed, giving place to sound initial 

 ideas of hygiene in the original compilation of the rock-work and 

 its drainage. 



P. alpina (Lois.) is a false name for P. viscosa. 



P. alpina (Salisb.) is a false name for P. auricula. 



P. alpina (Schleich.) is a false name for P.xpubescens. 



P. altaica of gardens is an altogether false name. It is usually 

 meant to apply to P. acaulis rubra ; whereas, in fact, it belonged 

 once in part, and unlawfully, to the type-form of the true P. amoena, 

 which is not in cultivation ; P. altaica has also been a m me for P. 

 Pallasii, Lehm. (a form of Oxlip), while it3 only tiue proprietor is a 

 variety of P. far,nosa, P. altaica, Lehmann. 



P. ambita is a form of P. obconica, and therefore quite unsuitable 

 to outdoor cultivation in Great Britain. 



109 



